Category Archives: Weather & conflict

It’s not the weather, it’s politics! Mike Davis’ “Late Victorian Holocausts” revisited.


Hsiang’s, Meng’s and Cane’s pompous chatter about El Niño and conflict made me want to read again one of those books they would certainly count among the “anecdotical” accounts: Mike Davis’ “Late Victorian Holocausts”. Far from being anecdotical this is an excellent, extremely well researched and documented historical study of the mismanagement of major droughts and floods in the 19th century India and China, colonies of the British Empire, and in Brazil, equally dependent on Great-Britain at that time, and the complex reasons that led to the probably biggest humanitarian disaster in the 19th century. The droughts and floods were caused by El Niño, all right, but the ensuing famines and hecatombs (Davis speaks of estimates between 12.2 and 29.3 Million for the two Indian famines of 1876-1879 and 1896-1902 alone) were caused by something rather different, much more complex and far more political than the weather: racist mismanagement, imperial arrogance, liberal free market ideologies, colonial interferences in local property rights, agriculture and rural production structures, administrative incompetence, “modernization” and its destruction of traditional patterns of solidarity and inter-communal help, abolishment of the state capacity in India and China where famines had been successfully prevented before…

It well takes the book’s 464 pages to expose the complex path from meteorological misfortunes to widespread famine and death. Yet, the core argument remains simple enough: Climatic vulnerability had always existed in those lands where the famines appeared in the late 19th century; however, it is only when weather hazards combine with the complete remodelling of the political and economic structures of these countries in the wake of their forced and violent “integration” into the world (or more precisely European and American) market that disastruous famines like these wiped out millions of people.

Three major effects of imperialism account for the heavy toll El Niño took in those years, as Davis painstakingly sets out in the third section of his book. “First, the forcible incorporation of smallholder production into commodity and financial circuits controlled from overseas tended to undermine traditional food security” (p. 289), “Second, the integration of millions of tropical cultivators into the world market during the late nineteenth century was accompanied by a dramatic deterioration in their terms of trade” (p. 290); “Third, formal and informal Victorian imperialism, backed up by the supernational automatism of the Gold Standard, confiscated local fiscal autonomy and impeded state-level developmental responses – especially investments in water conservancy and irrigation – that might have reduced vulnerability to climate shocks.” (p. 290).

Through a combination of military force and financial pressures agricultural production in these countries was redirected towards exports at the cost of local food security. Local governments (whether the provinces were under direct or indirect rule) were severed from own resources and capacities to undertake relief measures when disaster stroke. Davis tellingly compares the lack of response to the late 19th century famines with the relief campaign of the Qing Dynasty Governor-General Fang Guanchang who had established graneries all over the province of Zhili in Northern China and successfully combated a threatening famine after severe droughts in 1743-1744, shattering by the way common assumptions about the incapable and passive Chinese Qing state (see for more on Fang Guanchang and famine relief in imperial China, Liliane M. Li “Fighting Famine in North China”). Furthermore, as Davis points out in the gruesome first chapter of his book (which can be read online here:, a hideous melange of free market ideology (no state intervention!) and colonial racism offered convenient excuses for the colonial masters to do as little as possible for relief and to do the little they did as cynically and brutally as only those can do who do not consider people of other skin as humans. Nowhere is the obscenity of British imperial contempt more palpable than in Davis’ comparison of daily calorie rations distributed by the British in Indian “work-for-food” camps (where the starving were forced to heavy coolie labour) with the higher caloric value of food rations in Buchenwald (p. 39).

Davis is first of all interested in analysing the complex interworking of the three factors mentioned above but he does mention again and again how all these features that led to these enormous famines had already before created hardship and provoked armed rebellions such as the Taiping Rebellion, the Great Mutiny or the Boxer rebellion. In the context of analysing causes and reason for political violence, Davis’ comprehensive outline of the destruction the “world market” (which, again, was actually the British or other colonial motherland’s market) brought to India’s, China’s and Brazil’s economy and states points to three major issues that need consideration.

First, rather than poverty as such conflict analysis needs to consider the income land can generate for rural poor. This depends not solely on the productivity of land but also, and particularly in times of crisis, on external circumstances such as tax regimes, world food prices, food security capacities (e.g. irrigation and graneries) and rural finances, notably in terms of cash liquidity to come through times of failing harvests. Long before droughts or floods ruin harvests and income, peasants had fallen victim to vicicous circles of debt, forcing them out of the land or into cash crop production that seriously endangered their subsistence. Similarily, taxes need to be analysed whether they serve to support poor underclasses in times of crisis or, on the contrary, squeeze them to death.

Second, intimately linked but analytically an issue of its own, property rights and usage rights have to be carefully analysed to understand how, in the long term, poverty and exploitation are produced (or abolished). The greatest “innovation” brought by colonialism in the 19th century was the introduction of private property and the transformation of communal, shared land and other property forms into private property claims with the often accompanying dispossession of those who had before born the fruit of land usage. Far from being always the product of explicit violence or laws, dispossession has indeed taken a large variety of financial, economic and political forms.

Third, Davis’ fine analysis of market liberal discourses and “modernization” points to the urgent need of self-reflexive caution when it comes to analysing political and social conflict and struggle in strange societies. The currently so fashionable homo oeconomicus paradigm of conflict analysis (the famous rational utility-maximizing individual), a borrowing from liberal economics, often sprinkled with good doses of neo-Malthusianism, is entirely inappropriate in the light of the complex structural and relational dynamics through which 19th century imperialism has created those Victorian genocides. All the features of the 19th century liberalism are still well alive in global mainstream development politics, from the aversion against state intervention to the unbelievably efficient remodeling, through development assistance, peacebuilding and so-called statebuilding, of local economies to magically fit the business interests of Western and Northern economies rather than local needs.



It’s the weather, stupid! Hsiang, Meng and Cane on El Niño and conflict in “Nature”


Since a decade or so, large-N studies on the causes of war have appeared everywhere. They have this pretention of certainty around them, making believe that we could analyse wars like epidemics and find the cures through so-called “scientific” methods. But, well, yes,  uhm, hm, they might prove things but they are still a far cry from explaining anything as this recent high-profile publication shows.

One common claim of large-N studies is that their evidence goes beyond “anecdotical evidence” by which they obviously mean case studies, historical studies and other qualitative “stories”.  Solomon M. Hsiang’s, Kyle C. Meng’s and Mark A. Cane’s paper in the November 2011 “Nature” (online August 2011) are no exception to the rule. In their eye-catching article, the authors claim to provide a systematic evidence that changes in climate provide a higher risk of armed conflict. Their prove is a significant correlation of what they call “Armed Conflict Risk” with incidences of the El Niño Southern Oscillation (ENSO) affecting local weather. Yet, as commonly the case with over-aggregated, large-N studies the authors are incapable of offering an explanation for this phenomenon beyond highly speculative guesswork. So, their final concluding call is for more detailed, in-depth studies on how ENSO affects local conditions as to heighten the risk of war…to the anecdotes, isn’t it?

This paper epitomizes the fundamental problems of overaggregated, large-N studies. They look extremely sleek and fancy with their large data, colourful tables and complicated statistics but if someone would like to know more about the causes of armed conflict she will not receive an answer. At a closer look, when the glitter comes of, a number of serious questions arise over the methodology, the ontology of conflict and the political implications of such type of research. And most importantly, there remains a shallow taste of disappointment as we do not learn very much we haven’t known already from other studies, usually those diffamated as “anecdotical”.

If we look at the methodology – the alleged strong point of the essay – we can examine a bit closer the claim that the results reflect a global approach to the question whether weather and climate affect civil war risks. The authors argue that they can capture the global effect by grouping all countries over the world which are affected by ENSO climate changes. ENSO spreads indeed all over the world through so-called teleconnections. Yet, in order to construct one group of ENSO affected countries, we need to assume that these countries are affected by it in the same manner, for instance that it creates in all countries a particularly dry weather. However, this is not the case. There seems to be general agreement that teleconnections are non-linear and that El Niño-Southern Oscillation does not affect all regions in the same way.

Whereas in some regions ENSO creates a particularly dry and warm weather, other regions see a rise in precipitations and cold air. If the effects of ENSO on local weather is different in different parts of the world, then it becomes very unclear how we can compare these regions with each other beyond simply saying that their weather changes but it does not do so in the same way, so in the end we cannot know if it is the dry weather (droughts) or the rain (floods) that affect the countries in question (see for instance: Hoerling, Martin P., Arun Kumar, Min Zhong, 1997: El Niño, La Niña, and the Nonlinearity of Their Teleconnections. J. Climate, 10, 1769–1786.doi:<1769:ENOLNA>2.0.CO;2; Vera, Carolina, Gabriel Silvestri, Vicente Barros, Andrea Carril, 2004: Differences in El Niño Response over the Southern Hemisphere. J. Climate, 17, 1741–1753. doi:<1741:DIENRO>2.0.CO;2) . The analysis also does not adequately summarize the many possible differently interacting variables, at least not in this research where different effects of ENSO are lumped together in one group.

This is, of course, not very reassuring as to understanding what actually happens when particularly dry or particularly wet weather hit a country. The authors are, indeed, entirely at loss when they need to twist the correlation into an explanation. The authors present, in fact, a joly shopping list of possible causal mechanism (I absoluteley do not like the term “causal mechanism” as human society does not work like a machine but for the sake of simplicity I will use it here.) In their own words (p. 440):

  • “Generalizing our results to global climate changes other than ENSO will require an understanding of the mechanisms that link conflict to climate. ENSO has a proximate influence on a variety of climatological variables, each of which may plausibly influence how conflict-prone a society is. Precipitation, temperature, sunlight, humidity and ecological extremes can adversely influence both agrarian and non-agrarian economies.” Right! So, ENSO influences potato chip and micro-chip production in the same way? And with the same effects? How very interesting…
  • “In addition, ENSO variations affect natural disasters, such as tropical cyclones, and trigger disease outbreaks.” Well, well, and I always thought that diseases are caused by bacteries and viruses…
  • “All of these have adverse economic effects, such as loss of income or increasing food prices…” Which only matters if you spend more than 2/3 of your income on food and not on your iPhone or Gucci bag…
  • “Furthermore, altered environmental conditions stress the human psyche, sometimes leading to aggressive behaviour. We hypothesize that El Niño can simultaneously lead to any of these adverse economic and psychological effects, increasing the likelihood of conflict.” – of all these suppositions this one is certainly the most outstanding for its unabashed remininescence of good ol’ colonial malthusianist views of the savage…
  • “…and it is thought that economic shocks can generate civil conflict through a variety of pathways” – say, say…
  • “Furthermore, the influence of ENSO may exceed the sum influence of these individual pathways because it is a global-scale process that generates simultaneous and correlated conditions around the world.”

Well, isn’t that exactly the reason why we need micro-studies to know whether ENSO has impacted on cacao crops, potato and cabbage production, micro-chip production, financial markets, the inventiveness of Haliburton to sell their weapons or on the mood of parliamentarians and generals etc.? But if we would find that factor X has only been influenced indirectly by ENSO, let’s say that bad crops are only a problem because there are no other sources of income for peasants because their labour has been squeezed to starvation by extremely inequal landowning and exploitation structures? Wouldn’t that be exactly the kind of “anecdotical evidence” the authors think is not very useful for the study of the impact of climate on civil wars?